The Greatest Invention that Was Never Invented

Zsuzsanna Ardó reviews Guy Deutscher:
The Unfolding of Language,
Arrow Books, 2005

"The limits of my language mean the limits of my world."
Ludwig Wittgenstein, Tractatus Logico-Philosophicus

he Unfolding of Language sets out to be an entertaining historical evolutionary mystery tour and analysis of languages. It is about history but do not expect chronology: it investigates the general principles of the changes as they may have unfolded in the past. The relatively recent past. It starts the detective work from what the author calls the "me Tarzan" period—that is to say when lexemes already existed. So strictly speaking the book is not concerned with the unfolding of language per se, but more about the evolution of structures. It explores how grammatical building blocks developed, got streamlined overtime, how words got strung together, how the system became as endlessly subtle and complicated as it is today. Arguably, grammar is not half as sexy a word as language is, so it is not entirely surprising that grammar has not made it into the title of the book—even though, strictly speaking, the book is more concerned with the evolution of the grand structures, changing dynamics and self-amplifying complexity between words strung together i.e. structures usually thought of as grammar.

The evolutionary process of agglutination in the Indo-European languages and the vowel template in Semitic languages covered in the book do not include all language evolution, as the title might make you think. Some of the principles are demonstrated however, even if it does not aim to be comprehensive.

The subtitle of the book asserts that language is mankind's greatest invention. The term 'invention' in the subtitle is to grab your attention—but it is not to be taken seriously. The book itself extensively argues that language has in fact never been invented. "Language is mankind's greatest invention—except, of course, that it was never invented." (p.1)

The book feels like a 101 (intro) course for some (lucky) linguistics students at university, but one which has been rewritten to make it accessible for the general lay audience. The two audiences are, however, different in reality, and hence the book feels occassionally like two books not quite welded into one. The style of the book is—most of the time— light-hearted and entertaining on the one hand, with in-depth and quite heavy-duty readings on the other. Terminology is used sparingly and explained, but not always in the main body of the text. This means that lay readers would have to consult the glossary at the end of the book, which requires a different type of reading experience.

The narrative is journalistic, at times verging on thriller mode. Yet the book comes with a long list of academic references and further readings, essentially footnotes in the form of extensive endnotes. The postmodern variety of genres within the book can be quite enjoyable if the reader cottons on, rather than gets frustrated, that it calls for a non-linear, hyper-book reading style. The primary metaphor-network of the book is, not surprisingly, from the world of geology. It is argued with evocative and tactile imagery.

Conversation with Guy Deutscher

Guy Deutscher (GD): Well, 'unfolding' is not the same as 'emergence.' The book is not about the first emergence of words: I don't speculate about what happened when early hominids came down from the trees and started making their first grunts. I start from a stage when there were already words around, but language was still very primitive, with simple 'sentences' such as 'bring water' or 'mammoth die.' And from there, I try to show how complex language unfolds, or in other words, develops. I do this by looking at the way languages change today, and projecting that back onto the past. I talk about the 'structure' of language in the book, not about 'grammar,' because grammar often has the connotations of 'right' and 'wrong.' 'Grammar' for most people means the correct grammar they learnt in school. (Or the incorrect grammar they were given bad marks for.) But the structure of language has little to do with the often arbitrary norms of standard written languages. It is much more fundamental than that. And it is mastered by people who have never seen a grammar book in their life (or even any book at all).

ZA: The subtitle of the book asserts that language is mankind's greatest invention. Is it possible that the term 'invention' in the subtitle is misleading, considering that the book itself extensively argues that language has never been invented? As you say right at the beginning: "Language is mankind's greatest invention—except, of course, that it was never invented." (p.1) Why then the subtitle, one might muse.

GD: Well, the whole book revolves around the question of 'invention'—so it is quite fitting that 'invention' should appear in the title, don't you think? In one sense, language is our greatest invention, in that it is—I believe—our greatest cultural creation. But as I say on the very first page, this 'invention' was actually never consciously invented. Although the sophisticated structure of language looks like the work of a master architect, it must have somehow developed of its own accord. And the whole book is an attempt to explain how.

ZA: You seem to take it for granted that language is the 'greatest' invention (that never was invented) that makes us primarily human. Regarding our privileged position on the top of the food chain, how does language compare with the significance of tools? Art? Emotions? Are we not first and foremost psychological and social animals—and only then language users?

GD: Tools came before language—there is no question about that. Chimpanzees use tools, sometimes quite sophisticatedly, but they are not human. But I believe that practically everything we consider as distinctly human, including art, came from language. We cannot prove it, of course, because the people who made those cave paintings did not leave voice recordings behind them. But most scholars believe that the capacity for symbolic representation (Art) could only develop once the capacity of symbolic communication (language) was already in place. What's more, it seems clear that language and brain have co-evolved for a very long time. So even important elements of the genetics of our humanity are closely dependent on language.

ZA: The style of the book is remarkably light-hearted and entertaining on the one hand, with in-depth and quite heavy-duty readings on the other. Terminology is used at times but not always explained in the main body of the text. The narrative is journalistic, occassionally verging on thriller mode. Yet the book comes with a long list of academic references and further readings, essentially footnotes in the form of extensive endnotes. The postmodern variety of genres within the book can be enjoyed if the reader cottons on, rather than gets frustrated, that it calls for a non-linear, hyper-book reading style. You can dig deeper if you like but the whole of the narrative is generally not burdened with the footnotes or zillions of cross-references embedded in the text. How did this multi-genre and eclectic style approach 'unfold' as you were planning-writing-editing the book?

GD: I actually avoided jargon almost completely, and the extremely little terminology I did use is all explained - either in the text itself or in the glossary at the end. So I don't agree with you there. (Actually, a few linguists have criticized me for it—and claimed I should have introduced the readers to much more linguistic terminology.) But my aim was to make the book an interesting and entertaining read not just for academics, but for the general public. Still, I wanted to have scientific 'meat' in it, but I kept it mostly for the end-notes and the appendices. As for the 'postmodern variety of genres'—well, I think you refer mostly to the fact that one whole chapter is suddenly written as a sort of dialogue. I actually started writing it as a 'normal' text. But quite quickly, I realized that my argument in that chapter was essentially dialectic, and that it would be a sort of hidden dialogue anyway, in the form of, I say this, then someone objects that, etc. So I thought: if the argument is really a platonic dialogue, why try to hide it? Then I dressed it up a bit, of course. But essentially it was the content that dictated the form.

ZA How does the genre-style you opted for reflect the kind of audience you had in mind as you were writing? The book feels like a 101 (intro) course for some (lucky) linguistics students at university, but one which has been rewritten to make it enjoyable for the general lay audience. Would you agree that the two audiences are, however, different in reality, and hence the book feels like two books not quite welded into one?

GD: Indeed, the two audiences are different, and while I want both students and the general public to enjoy it, my priority was ultimately with the general public (hence not introducing much terminology). My idea was that students could 'cope' with the lack of complex jargon, whereas the general audience would just give up if there was much of it. Ultimately, the 'idealized reader' I had in mind was myself at the age of 18. I was trying to answer many of the questions that fascinated me then, and which ultimately made me go and study linguistics.

ZA: The primary metaphor-network of the book is, not surprisingly, from the world of geology. It is argued with evocative and very physical, tactile imagery. Terms like erosion, devastation, decay, destruction crop up again and again. This means though that the book constructs language change as a negative, destruction narrative—only to turn it around at the very end, Hollywood movie-style. We come a full cycle in the end: creation comes from destruction itself. But is it possible that by this time the reader is saturated with the negative terms of destruction? Could it not be argued that language change could be framed in terms of streamlining rather than destruction and devastation?

GD: There is quite a long tradition in linguistics of borrowing concepts and images from geology. In fact, one of the most important pillars of modern linguistics, as I say in the book, is the idea that 'the present is the key to the past,' which is of course originally a geological insight, used to explain the features of the earth today not by some divine catastrophic events, but by very slow processes occurring over a very long period of time.

As for 'streamlining' - I would say that 'streamlining' is the result of the combination of the forces of destruction and the forces of creation. I tried to show in the book that 'erosion' (caused e.g. by sloppy pronunciation) can bring about changes that are very different from mere 'decay.' Erosion is not only a negative influence on language, which tears away and rips apart existing structures. In combination with other forces of change (e.g. our natural expressive urge to pile up of words to make 'stronger' phrases), erosion is also a regenerative force that constantly creates new and leaner structures from overweight multi-word phrases. Erosion is a highly useful compacting mechanism which allows us to convey ideas faster and more efficiently. Erosion checks the excesses of expressiveness, just as expressiveness repairs the excesses of erosion.

ZA: You argue that only small, isolated, primitive societies kept their fossilized languages. However, Hungarian for example, is neither small, nor is it primitive or isolated. Quite the contrary. Hungarian has incorporated a wide variety of influences in abundance yet it survived in its relatively fossilized form right in the busy cross-roads of Central Europe. How revealing is Hungarian (and other Finno-Ugrian languages) about language change in terms of having survived in a relatively fossilized state in the midst of rapid and constant language change around them? You argue that it is exactly pattern-defying examples that give us insight into the past. Are languages such as Hungarian then the proverbial Popperian white ravens one is looking for as a researcher?

GD: It is true that Hungarian has remained as a separate language, and has not 'drowned' in the languages around it. However, Hungarian has by no means remained isolated. It shows substantial influence (borrowed words, for instance, as well as convergent grammatical features) from many languages in the region, from German to Turkish. But again, that is by no means unusual. Practically every other language in Europe has developed over the last millennia under influence of its neighbouring languages. The only difference is that Hungarian is not Indo-European, that is, originally they came from a different ancestor language from the rest of the languages of Europe.

ZA: Is there correlation between the degree of agglutination preserved in a language (e.g. in the Finno-Ugrian ones) and its degree of fossilization in as far as it gives insight into early stages of development?

GD: Well, there are different questions involved here, all very complex. One is: do all languages change roughly at the same rate? Or are some much more conservative. And if there are different rates of change, what is responsible for that. We have no simple answer. But it seems that contact certainly helps changes, and that intense contact results in quick changes (e.g. as happened in England after 1066).

ZA: You state on p 25 that "even in today's languages, it is fair to say that arranging the bricks in a particular order is still the most important element in the art of sentence construction." Compare this statement with Hungarian for example, where you can be almost as free as you like with the word order, to emphasize subtle shades of meanings inherent in a sentence. Here, you cannot really go wrong, unlike, for example, in English where it can make a slight difference whether you say "The dog bit the man." or "The man bit the dog." in the classic example

You are, for example, perfectly free to rearrange the word order in many different mutations in the sentence below, for the same sentence in English.

GD: Of course, some languages like Hungarian are much freeER in word order than English. I show similar things with Russian and modern Aramaic on pp. 33-34. And there are some Australian aboriginal languages which are even freer than Hungarian, Russian, Latin, and so on. BUT—no language is entirely free in word order. The fundamental word order principles I talk about on pp. 214-224 of the book are at the basis of all languages, without which communication would just collapse.

Just as one example, there is no language where the Caesar's phrase 'veni vidi vici' could just be juggled without changing the meaning of the sentence. The whole point about it is that the order of words here reflects the order of events in reality: *first* I came, *then* I saw, and *finally* I conquered. Of course, you can add a word to the sentence and say 'BEFORE I conquered, I came and saw'. But there is no single language where you can simply juggle the words around without severe consequences for the meaning. This is one basic principle of word order that is called 'time iconicity'. But there are others, as I show. Word order is never entirely free.

ZA: A biological physicist reader of the book I was discussing it with was impressed by the scientific evidence in the book but came away with the question: does the evolutionary process of agglutination in the Indo-European languages and the vowel template in Semitic languages cover all language evolution?

GD: The processes I demonstrated in Indo-European and Semitic are indeed universal, and in that sense cover the whole evolution of language, not just that of Indo-European and Semitic. But that does not mean I covered every single aspect of linguistic structure. For example, I said no single word about tone and tonal languages, where different tones or intonations are used to differentiate between words. (This happens in East Asian languages, African languages, but also to some extent in Swedish and Norwegian). So no, I didn't talk about everything. But what I was saying is that I've tried to show the basic principles, and in that way argue that the rest of the details could be filled along similar lines.

ZA: The coda of your book mentions that out of the approximately 6000 languages—although this varies according to sources—one dies bi-weekly. Firstly, how are these numbers arrived at? Secondly, do you find this as an ultimately down-ending narrative?

GD: The figure of 'two languages dying every month' is something that has become a standard phrase among linguists, but, as you rightly sense, it is ultimately based on very vague estimates. There is no other way: after all, there are even languages dying today which we don't even know exist! In my book, this question is mentioned only in passing in the very end, so I don't go into the details.

ZA: The issue of language death is dealt with in the literature elsewhere, but I am wondering if you'd care to sum up your view as to how the death rate is arrived at, and what the real implication of this evolutionary process is. If a language dies bi-weekly it means, if the statement is as unqualified as it is, that the number of languages in the end will be reduced to a handful. So my question on dialect vs language with army and navy was actually meant to be not about proliferation of languages (e.g. Serbo-Croatian splitting into two), but the opposite end of the spectrum i.e. the world of languages (hence armies and navies) reduced to a couple or just one.

GD: Well, *if* it is indeed 2 languages a month, that means around 24 languages a year, and so 2400 languages in the next 100 years. There are estimated to be about 6000 languages in the world today, so that would still leave more than half of them. However, this figure is an extremely vague estimate, and possibly quite a low one. In fact, I think this estimate was derived 'backwards' - not from knowing how many languages are dying every month, but by estimating, globally, how many languages will die within a generation or two, and then dividing it up to years/months etc. The global estimates are generally based on factors such as size of the population, level of existing bilingualism, and level of retention in the new generation (there are many small languages which are still spoken by a fair amount of people (a few hundred, or even more), but these people are old, and younger people have already switched to another language. This is a certain indication of imminent language death).

ZA: The fascinating question not raised in the book but well worth asking is this: where is this paradigm pointing to?

GD: This question has both political and scientific dimensions. The political dimension is rather vexed: on the one hand, there are many people who think (and say) that we will all be better off if there are fewer languages around, since it will be easier to communicate. We are, in effect, undoing Babel. I don't agree with these people, but I can understand their point of view. My own opinion is that bilingualism or multilingualism is an equally effective way of facilitating global communication, but much less corrosive of other things that many people hold dear. After all, life is not just about economic success. Language is the carrier of a people's culture, and when a language dies, people often feel robbed of their identity. (And if they don't, then their children or grandchildren will, as we know, for instance, from the thousands of Native Americans who are today desperately trying to relearn the languages which their parents or grandparents forsook.) The scientific dimension is grim but fairly simple: we are irrevocably losing a vast resource of knowledge, about language itself, about human culture, and about the relation between them. The little we can do is to try and document as many languages as we can, as thoroughly and reliably as we can, before they are forgotten. But it's fairly hopeless race, because documenting even one language properly takes years of intense effort.

ZA: If languages have such a dizzying death-rate, is it only a question of time when the thousands of languages will be reduced to only a few—indeed, only one perhaps? If so, which one(s) will it (they) be? And, perhaps more importantly, why is this the case in your view?

GD: I don't think we will get just one language in the foreseeable future, not even a handful. But it's not a great secret which languages will go and which will stay: it's ultimately a question of numbers. The languages that will go are overwhelmingly those of small societies, of 'simple' unindustrialized peoples, with rich oral tradition but no writing, so there will be so little so show for themselves. The safest languages are the ones with the largest number of speakers. Perversely, at least from the point of view of linguistics, it would have been much better if English and the major European languages were to disappear instead. We know too much about them already...

ZA: How will this scenario impact on culture, identity, society, literature and so on? What are the political implications and the power dynamics implicit in this forecast? Who will be able to read the libraries of books in various languages if in a couple of generations we will be speaking just one language?

GD: Well, I don't think we would all be speaking just one language in a hundred years time. Not even two. And the overwhelming majority of languages that will die are unwritten ones. So in that respects, libraries are not critically threatened. Having said this, even today, scientific literature is becoming predominantly monolingual: in English. The rich scientific tradition in languages such as German or French, for instance, has become largely inaccessible to a large number of academics in the Anglo-Saxon world (not to mention the general public).

ZA: In the wider context of the above, that is to say the trend of rapidly vanishing languages, what is your view of the statement, credited to the linguist, Max Weinreich, from the opening session of the 19th Annual YIVO Conference in New York City in 1945, that "a language is a dialect with an army and a navy"?

GD: He (I think he claimed it was actually someone in the audience in one of his lectures) got it spot on. There are no hard and fast rules for determining when two 'dialects' become different languages (as we see with Serbian and Croatian, which were 'dialects' before the war in former Yugoslavia, but different languages after). It ultimately boils down to what the speakers feel. And those with an army and a navy are more likely to feel they are speaking an independent language.